Small Target Detection And Tracking Based On Temporal Predictions Of Background In Infrared Image Sequences

It is important to research the algorithms of small target detection and tracking because the system of detecting and tracking target using infrared thermal imaging instruments is the chief component of those precise control and guide weapens while the algorithms are the core of this system.The general situation of the technique of small target detection and tracking is summarized in this thesis. Following the summary, the problem of detecting and tracking small, moving, low amplitude objects in infrared image sequences that also contain moving nuisance objects and background noise is addressed in the thesis.First of all, the idea of temporal predictions of background is put forward to solve the above problem. The small target is detected by taking full advantage of the information of background pixels. And the algorithm of small target detection is brought forward based on temporal predictions of background in infrared image sequences.Secondly, in the following tracking process, the design of system software needs the small target to appear in the designed watching field and the information of continuity in tracking process is less than that in detecting process because the background in watching field is changing. Thus, the small target is tracked according to the information of distribution, gray and shape of the small target and the corresponding background. And the algorithm of small target tracking is put forward based on multi-feature fusion of the target and the corresponding background in infrared image sequences.Finally, the effectiveness of these algorithms mentioned above are demonstrated by applying the algorithms to real world infrared (IR) image sequences containing targets of opportunity. At the end of this thesis the main contribution of the research is summerized and some promising research directions in the field of small target detection and tracking are proposed.